The Human+AI Equation: The Missing Yardstick

In 2025, one major technology firm cut close to a fifth of its global workforce despite a sharp increase in net income that same year. Another, posting billions in quarterly profit, eliminated thousands of roles to strip out management layers. Announcements like these have become routine, and they tend to share a vocabulary about restructuring, efficiency, and reinvestment toward AI. The language is confident and forward-looking while skipping past an uncomfortable fact: most organizations do not measure the things that would justify any of it.

AI Didn’t Create the Measurement Problem

Across this Human+AI Equation series, we have dissected three proxies that have long stood in for valuable knowledge work. We showed how time saved is not the same as value created; how visible effort rewards the appearance of work over its result; and how a volume of ideas is not the same as turning a useful one into reality. Each of these proxies is a shadow of value, not the thing itself. GenAI has broken all three at once, exposing how wide the gap always was between these proxies and the value they stood for.  

The explanation that recent layoffs are due to pandemic-era over-hiring only underscores the point. If a company can comfortably carry thousands of roles for years that it suddenly deems unnecessary, it was never sizing its workforce by delivered value to begin with. Few organizations have reliable ways to track operational complexity or the ratio of genuine value to headcount cost. What many of these layoff announcements describe is not the result of strategy but a reasonable-sounding explanation that follows a decision made largely on gut instincts.  

Executives are responding to market pressure to signal AI transformation, investor expectations for efficiency, and internal pressure to reduce cost quickly—which is often easier than measuring value. Many decisions are not uninformed so much as pressured. Cost is measurable. Value is not.

When Cost Is Visible, and Value Is Not

Rather than an accusation of bad faith by companies, this is an observation about how organizations define and measure the value of knowledge work. When an organization cannot accurately quantify the value of its human workforce, it cannot reliably tell a redundant role from one whose value was never articulated. The results are, ironically, a lot of wasteful inefficiency.

In one 2026 survey of HR leaders whose organizations had cut roles to make way for AI, about two-thirds admitted they had rehired for some of the roles they had eliminated. A reversal on this scale is not a sign that conditions have changed, but that the original decision was made without a clear view of which roles really needed to be cut. Rehiring doesn’t just signal inefficiency; it is evidence that organizations lacked a reliable way to distinguish value in the first place.

The Problem With AI Replacement Claims

“This role could be done mostly by AI” may seem like a reasonable statement on the surface. However, without any prior measurement to baseline against—including a bar for the quality or usefulness of output—it is difficult to determine the ROI of a one-to-one human role replaced by AI. Current claims of AI substitution are subjective by default because many white-collar roles may be undergoing their first serious appraisal of business value.

To be clear, this is also not a failure of workers to upskill, anticipate, or avoid roles destined for AI replacement. In fact, that is difficult to do when most organizations create new roles in ways that are largely divorced from known business value. Headcount is planned not just against business justification (itself an imperfect proxy of real value) but also against budgets, manager overload, and team size. Roles are often backfilled by default when someone leaves.

Standard talent-acquisition metrics such as time-to-fill and cost-per-hire measure the logistics of staffing, not the value a role adds. A role can enter the organization without a value premise, be managed for years without one, and ultimately be cut without one as well. AI did not create this blind spot; it simply provided a commonly defensible audit of role value.  

Further, two people in the same role may produce radically different value, as AI amplifies individual differences in leverage and impact. This is a discussion point for another piece, but it is worth keeping in mind. For now, even ignoring individual differences in value creation and focusing only on role differences is compelling enough.

What Work Actually Creates Value?

The real dilemma facing executives in the next phase of AI-enabled knowledge work includes: understanding which work, and which roles, are critical for value creation.

Some roles look low value on short-term financials but are critical to long-term capability, such as R&D or leadership development.  Developing this understanding is not a new discipline; it is the same one organizations already apply to every other part of the business they take seriously.

It means defining the outcome a role exists to produce, then finding a defensible way to instrument that outcome directly, rather than tracking the activity merely correlated with it. Measuring value in knowledge work is inherently difficult because outcomes are delayed, collective, and often uncertain. Difficulty, however, does not justify defaulting to weak proxies.  It is the same logic behind any sound measure: be explicit about the construct you care about, choose indicators that track it rather than its proxies, and stay willing to ask whether your measure and your intent have quietly drifted apart.

Four Questions Every Organization Should Be Able to Answer

Organizations that begin to measure value-driven work will be best positioned to retain, reward, and develop their most critical talent in the confusing landscape of AI enablement of knowledge work.

There is a simple diagnostic for any leader who wants to know where their organization stands. Ask these four questions—and notice how hard it is to answer honestly:

1. For a given role, can you state what it is meant to produce in terms of outcomes to the business, not job activities, in a way that someone outside the function would recognize as valuable?
2. If the role suddenly disappeared tomorrow, what specifically would degrade about the business, and how would you know?
3. The last time a significant number of roles were cut, what evidence was used to separate the ones you kept from the ones you let go?
4. For the roles you currently consider most critical, is that judgment based on something measured or something intangible?

To be sure, deliberate evaluation of the value produced by roles does not eliminate the need for layoffs or restructuring. Some roles will genuinely need to be cut as businesses and technology grapple with technological change and economic volatility.

However, a company that can reliably show which roles are being cut based on measurable outcomes and strategically chosen trade-offs will be able to manage valuable talent with greater precision, fairness, and credibility. Doing so requires more of a shift in mindset than a shift in process. The same talent-acquisition processes that manage hiring today can be organized around the questions above, rather than logistics.

The Real Human+AI Question

We close our Human+AI series by stating plainly that this equation was never only a question of what humans can do versus what AI can do. It is instead a question of whether organizations are willing to take a deeper look at their own value-producing work to redesign around new technologies, including AI.

For decades, organizations could not, and did not have to, directly measure the value of knowledge work because the imperfect proxy measures held up well enough, and it was a given that human workers were the primary agents by which things got done. As AI brings about a new potential reality, we can no longer ignore the question of how we will measure the true value of human work.

The organizations that emerge strongly in the Human + AI era will be the ones that stop measuring the shadow of valuable work and start measuring the value itself.  Without a credible yardstick, organizations will continue to mistake cost reduction for value creation, and AI will only accelerate that confusion.  

‍

The Human+AI Equation series explores why human judgment, attention, and commitment matter more—not less—in a GenAI enabled world. Read the other articles in the series: Turning Time Savings into Real Value, Letting Go of the Effort Illusion, and Escaping the Idea Trap.  

Learn more about Korn Ferry’s Human + AI work and how organizations are redesigning roles, workflows, and talent strategies for an AI-enabled future.